Phantom? Don't worry about it.
The "open" switch must block the open circuit voltage, 48V. The arcing voltage between round electrodes in air is 300V; even if you get sharp points on the contacts 48V won't jump across.
The closed switch must pass the current. A 100 Ampere contact is a large fraction of an inch. A 10mA contact is 8000 times smaller.... basically the smallest contact which is economic to stamp and assemble will handle 10mA. If it has a handle, it will have big-enough contacts.
Other important considerations:
How many times? If you flip 10 times a second 24/7, you need ample reseve for mechanical wear.
Can there be a "fault"? My house main breaker is rated to break 100A as needed; it is also rated to break 10,000A _once_ without exploding, because a heavy dead-short in the box could suck 10,000A from the street (not MY street, but well-wired homes). In this case your 3K4 Phantom resistors are "permanently" in-line, and limit current to 14mA worst-case, just as the 0.024 ohms of 40 feet of feeder-wire limit house-faults to 10,000A.
If it fails, does anybody die? Big industrial machinery MUST switch correctly. I like my house breaker to be over-rated because if a fire started while I was out the dogs would die. The downside of a failed Phantom is less.
Are we switching nasty loads? Worst-case is large current through a large inductor. When you break, the coil voltage rises toward infinity until the switch breaks-over. On DC supply, ALL the stored energy must be dissipated in this arc. On AC supply there is a tendency for the arc to stop in a half-cycle. (Which is why power tools with AC/DC motors often have AC-only ratings: the switch was not bought to tolerate constant DC breaks.) Switching a large cap onto a voltage is less tough, but is is wise to throw 100 ohms between switch and any 100uFd cap you may have put on the switched side.
> rated for 30VDC
24V-30V falls under different regulations. It is unlikely to electrocute; and for low current it is unlikely to start a fire. (12V 200A car wiring is different and sure can start a fire.) If they claim over 30V, they may have to do many more tests.
Farnell's stupid site shows switches rated to 1mA. That's crazy. Especially for Illuminated switches: the lamp must be taking more than 1mA, and why should the load contacts be weaker?
While I might avoid the 1mA part, any of the 10mA and up will be fine.
> For line level audio signals, chose switches with a lower voltage rating.
Actually, avoid cheap high-power switches. I used 250V 10A toggles to patch coax Ethernet; they were user-rugged. They failed, open. High-power switches use contacts which tarnish. If you switch a 100 watt lamp or 120V 7A motor, the load-power burns off the tarnish faster than it grows. 2 years in a nasty office rarely switching 1V 20mA, they fail. In my network, the whole net was down. In your Phantom, worst is that you have to move to another channel and replace the switch later.